[Bug c++/54170] New: Call to lambda elided

[Bug c++/54170] New: Call to lambda elided

[bekenn at yopmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

             Bug #: 54170
           Summary: Call to lambda elided
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: bekenn@yopmail.com


Created attachment 27935
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27935
preprocessed source

gcc version: 4.7.0
target: mingw32
Configured with: ../gcc-4.7.0/configure
--enable-languages=c,c++,ada,fortran,objc,obj-c++ --disable-sjlj-exceptions
--with-dwarf2 --enable-shared --enable-libgomp --disable-win32-registry
--enable-libstdcxx-debug --disable-build-poststage1-with-cxx
--enable-version-specific-runtime-libs --build=mingw32 --prefix=/mingw
command line: gcc -v -save-temps -std=gnu++0x -c -g src\lambdatest.cpp
COLLECT_GCC_OPTIONS='-v' '-save-temps' '-std=gnu++11' '-c' '-g' '-mtune=i386'
'-march=i386'

lambdatest.cpp:
#include <utility> // for std::forward

// externally defined
void ExternalFunction1(const char*, int);
void ExternalFunction2(int, void**);

template <class Callable, typename... ArgTypes>
void* Call(Callable native_func, ArgTypes&&... args) noexcept
{
    return native_func(std::forward<ArgTypes>(args)...);
}

void* test_lambda(void* functionData, int argc, void* argv[])
{
    return Call([=]
    {
        ExternalFunction1("Lalala...", 0);
        ExternalFunction2(argc, argv);
        return nullptr;
    });
}


---

This compiles without error messages, but gives incorrect output.  The call to
native_func on line 10 is left out, and Call simply returns 0 instead.  This
can be verified by examining the generated assembly:

__Z4CallIZ11test_lambdaPviPS0_EUlvE_IEES0_T_DpOT0_:
    pushl    %ebp
    movl    %esp, %ebp
    movl    $0, %eax
    popl    %ebp
    ret

No code is generated for the lambda body, and the resulting object file does
not contain references to ExternalFunction1 or ExternalFunction2.

[Bug c++/54170] Call to lambda elided

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
   Last reconfirmed|                            |2012-08-04
     Ever Confirmed|0                           |1

--- Comment #1 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-08-04 00:07:33 UTC ---
Can you please add a self contained testcase, not requiring assembly
inspection? Thanks.

[Bug c++/54170] Call to lambda elided

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #2 from Jonathan Wakely <redi at gcc dot gnu.org> 2012-08-04 10:47:49 UTC ---
Dup of PR 52988?

[Bug c++/54170] Call to lambda elided

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-08-04 11:35:34 UTC ---
Indeed, I suspected that immediately. But without a self contained testcase not
requiring inspection of the assembly personally I'm not 100% sure (and I'm in
the middle of 2-3 other things).

[Bug c++/54170] Call to lambda elided

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #4 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-08-04 11:37:12 UTC ---
If submitter can try current 4_7-branch or mainline, it would be easy for him
to double check.

[Bug c++/54170] Call to lambda elided

[bekenn at yopmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #5 from bekenn at yopmail dot com 2012-08-06 04:44:17 UTC ---
Created attachment 27948
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27948
Full application showing issue.

Sorry for the wait.  Here is a complete program showing the issue.  Compiled
with:
    g++ -std=gnu++0x -g lambdatest.cpp -o lambdatest.exe

Traces have been added to show the current position of execution.

Expected output:
start
in test_lambda
in Call
in lambda
Adding 5
total: 5
after first call
in test_lambda
in Call
in lambda
Adding 20
total: 25
after second call
in test_lambda
in Call
in lambda
Adding -256
total: -231
after third call

Actual output:
start
in test_lambda
in Call
after first call
in test_lambda
in Call
after second call
in test_lambda
in Call
after third call

[Bug c++/54170] Call to lambda elided

[bekenn at yopmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

bekenn at yopmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #27935|0                           |1
        is obsolete|                            |

--- Comment #6 from bekenn at yopmail dot com 2012-08-06 04:46:01 UTC ---
Created attachment 27949
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27949
preprocessed source for new example

[Bug c++/54170] Call to lambda elided

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW
                 CC|                            |jason at gcc dot gnu.org

--- Comment #7 from Jonathan Wakely <redi at gcc dot gnu.org> 2012-08-06 07:51:48 UTC ---
hmm, it still fails on trunk.

Jason, the fix for PR 52988 seems to be incomplete.

[Bug c++/54170] Call to lambda elided

[bekenn at yopmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #8 from bekenn at yopmail dot com 2012-08-06 08:11:38 UTC ---
(In reply to comment #4)
> If submitter can try current 4_7-branch or mainline, it would be easy for him
> to double check.

I installed mint in a virtual machine to test those, but I couldn't get either
trunk or gcc-4_7 to build; I got these errors, which I haven't bothered to
investigate:

build/gengtype.o: In function `create_optional_field_':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype.c:853: undefined
reference to `lexer_line'
build/gengtype.o: In function `adjust_field_rtx_def':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype.c:989: undefined
reference to `lexer_line'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype.c:1045:
undefined reference to `lexer_line'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype.c:1055:
undefined reference to `lexer_line'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype.c:1121:
undefined reference to `lexer_line'
build/gengtype.o:/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype.c:1160:
more undefined references to `lexer_line' follow
build/gengtype-parse.o: In function `token':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:53:
undefined reference to `yylex'
build/gengtype-parse.o: In function `parse_error':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:142:
undefined reference to `lexer_line'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:142:
undefined reference to `lexer_line'
build/gengtype-parse.o: In function `struct_field_seq':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:692:
undefined reference to `lexer_line'
build/gengtype-parse.o: In function `type':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:726:
undefined reference to `lexer_line'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:758:
undefined reference to `lexer_line'
build/gengtype-parse.o:/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:758:
more undefined references to `lexer_line' follow
build/gengtype-parse.o: In function `parse_file':
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:951:
undefined reference to `yybegin'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:986:
undefined reference to `lexer_toplevel_done'
/home/bekenn/Projects/gcc-build/gcc/../../gcc-4_7/gcc/gengtype-parse.c:991:
undefined reference to `yyend'
collect2: ld returned 1 exit status

[Bug c++/54170] Call to lambda elided

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #9 from Jonathan Wakely <redi at gcc dot gnu.org> 2012-08-06 08:32:28 UTC ---
I think those errors indicate you don't have Flex installed

[Bug c++/54170] Call to lambda elided

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at gcc dot       |paolo.carlini at oracle dot
                   |gnu.org                     |com

--- Comment #10 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-11-30 12:31:15 UTC ---
It is really an incomplete fix for PR52988, changing nullptr to (void*)0 works
around the problem. Looking a bit into it.

[Bug c++/54170] Call to lambda elided

[paolo at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #11 from paolo at gcc dot gnu.org <paolo at gcc dot gnu.org> 2012-12-03 16:01:43 UTC ---
Author: paolo
Date: Mon Dec  3 16:01:32 2012
New Revision: 194098

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=194098
Log:
/cp
2012-12-03  Paolo Carlini  <paolo.carlini@oracle.com>

    PR c++/54170
    * cvt.c (cp_convert_to_pointer): Don't discard side-effects from
    expressions of nullptr_t.
    * typeck.c (build_ptrmemfunc): Likewise.

/testsuite
2012-12-03  Paolo Carlini  <paolo.carlini@oracle.com>

    PR c++/54170
    * g++.dg/cpp0x/lambda/lambda-nullptr.C: New.

Added:
    trunk/gcc/testsuite/g++.dg/cpp0x/lambda/lambda-nullptr.C
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/cvt.c
    trunk/gcc/cp/typeck.c
    trunk/gcc/testsuite/ChangeLog

[Bug c++/54170] Call to lambda elided

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.8.0

--- Comment #12 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-12-03 16:10:13 UTC ---
Fixed for 4.8.0.

[Bug c++/54170] Call to lambda elided

[foom at fuhm dot net]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

James Y Knight <foom at fuhm dot net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |foom at fuhm dot net

--- Comment #13 from James Y Knight <foom at fuhm dot net> ---
Possibly should be backported to 4.7 branch?

[Bug c++/54170] Call to lambda elided

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

--- Comment #14 from Jason Merrill <jason at gcc dot gnu.org> ---
The patch does seem safe enough for 4.7, even though I usually prefer not to
backport C++11 fixes to older release series.

[Bug c++/54170] Call to lambda elided

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.8.0                       |4.7.4

--- Comment #15 from Jason Merrill <jason at gcc dot gnu.org> ---
Fixed for 4.7.4.

[Bug c++/54170] Call to lambda elided

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=54170

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |fox at ucw dot cz

--- Comment #16 from Jonathan Wakely <redi at gcc dot gnu.org> ---
*** Bug 61302 has been marked as a duplicate of this bug. ***